Data Spaghetti (Thursday session)

On Thursday I did a few sessions experimenting with the RIM software. I found that the Stroke Efficiency metric is really sensitive to the calibration.

I did an experiment where I recorded the entire session in 4 parts. Between parts I took the phone out of the holder while not running RIM, waved with it, and put it back in the holder. I believe the only difference was the time I waited while not rowing but RIM already running.

I saw quite big differences in the Stroke Efficiency metric. In the first 2 “parts”, I saw values around 0.4 maximum, and even negative values. In the last two “parts” I saw values above 2 at 20spm. I think these were the “real” values.

This morning I managed to grab the analysis data from the analytics.rowinginmotion.com website using jsfiddle.net and read them into pylab. Here’s an example:

This is just a raw plot of the acceleration data of a few strokes at 30spm. Now it will be easy to compare my rowing model with real data!

The other application is CrewNerd. Also CrewNerd can export acceleration data, but I have yet to experiment with it. Will do in the near future.

But I took a different look at some of my 4x1km data this morning. I exported all 4 intervals into CSV tables, dumped all the data in one table and started playing with pivot charts:

I still have to find a useful plot that enables to draw conclusions from the data. Here is Check vs Speed for different stroke rates:

Data Spaghetti!

No training today. I feel a cold is coming on me, and I will race tomorrow.